Session G5: Cardiovascular VI

Disease severity index derived from hemolysis evaluation

Several cardiovascular diseases (CVDs) are characterized by stenosis of the vessel, leaflet malfunction, disturbance of blood flow (vorticity) due to geometric deformation or abnormal growth, and development of jet flow due to ventricle overload. All of these abnormalities are followed by degeneration of inner wall of the heart and the arteries and red blood cell damage (hemolysis). In this study, identification and classification of CVDs are being performed based on hemolysis evaluation (HE). Two commonly used hemolysis models are implemented to our computational fluid dynamics simulations of CV system. The capability of HE on disease diagnosis is investigated. The analysis will be carried out on our CVD templates such as artery stenosis or pulmonary artery hypertension. HEs depend mainly on the strain rate and for some computational hemolysis models there is a threshold of strain of which the hemolysis will not take place. In the current study, we investigate the effect of thresholding besides using pseudo exposure time for steady state simulations on the blood damage evaluations. Details of our methodology for HE by post processing simulation results without necessity of re-running the simulations will be presented.   

Nonlinear flow response of soft hair beds

We are hairy inside: beds of passive fibers anchored to a surface and immersed in fluids are prevalent in many biological systems, including intestines, tongues, and blood vessels. Such hairs are soft enough to deform in response to stresses from fluid flows. Fluid stresses are in turn affected by hair deformation, leading to a coupled elastoviscous problem which is poorly understood. Here we investigate a biomimetic model system of elastomer hair beds subject to shear- driven Stokes flows. We characterize this system with a theoretical model which accounts for the large-deformation flow response of hair beds. Hair bending results in a drag-reducing nonlinearity because the hair tip lowers toward the base, widening the gap through which fluid flows. When hairs are cantilevered at an angle subnormal to the surface, flow against the grain bends hairs away from the base, narrowing the gap. The flow response of angled hair beds is axially asymmetric and amounts to a rectification nonlinearity. We identify an elastoviscous parameter which controls nonlinear behavior. Our study raises the hypothesis that biological hairy surfaces function to reduce fluid drag. Furthermore, angled hairs may be incorporated in the design of integrated microfluidic components, such as diodes and pumps.   

Secondary flow structures in a 180$^\circ$ elastic curved vessel with torsion under steady and pulsatile inflow conditions

Secondary flow vortical structures were investigated in an elastic 180$^\circ$ curved pipe with and without torsion under steady and pulsatile flow using particle image velocimetry (PIV). The elastic thin-walled curved pipes were constructed using Sylgard 184, and inserted into a bath of refractive index matched fluid to perform PIV. A vortex identification method was employed to identify various vortical structures in the flow. The secondary flow structures in the planar compliant model with dilatation of 0.61\%-3.23\% under pulsatile flow rate were compared with the rigid vessel model results, and it was found that local vessel compliance has a negligible effect on secondary flow morphology. The secondary flow structures were found to be more sensitive to out of plane curvature (torsion) than to vessel compliance. Torsion distorts the symmetry of secondary flow and results in more complex vortical structures in both steady and pulsatile flows. In high Re number steady flow with torsion, a single dominant vortical structure can be detected at the middle of the 90$^\circ$ cross section. In pulsatile flow with torsion, the split-Dean and Lyne-type vortices with same rotation direction originating from opposite sides of the cross section tend to merge together.   

Hemodynamics of Aortic Stenosis and Implications for Non-invasive Diagnosis via Auscultation

Aortic stenosis refers to the abnormal narrowing of the aortic valve and it is one of the most common valvular diseases. It is also known to generate ejection murmurs, which contain valuable disease-related information. However, an incomplete understanding of the flow mechanism(s) responsible for the murmur generation, as well as the effect of intervening tissue on murmur propagation has limited the diagnostic information can be extracted through cardiac auscultation. In this study, a canonical model of the aorta with stenosis is used, and a multiphysics computational modeling approach is employed to investigate the generation and propagation of the murmurs. First, direct numerical simulation (DNS) is used to explore the hemodynamics of the post-stenotic flow. Then, a high-order, linear viscoelastic wave solver is used to investigate the wave propagation in a modeled thorax. The results show that both the aortic jet and the secondary flow contribute significantly to the murmur generation. The murmur signals on the epidermal surface are measured and analyzed. The break frequencies obtained from the spectra of cases with different degrees of stenosis are found to follow a universal scaling. The implications of these results for cardiac auscultation are discussed.   

Correlation between vortices and wall shear stress in a curved artery model under pulsatile flow conditions

One of the most physiologically relevant factors within the cardiovascular system is the wall shear stress. The wall shear stress affects endothelial cells via mechanotransduction and atherosclerotic regions are strongly correlated with curvature and branching in the human vasculature, where the shear stress is both oscillatory and multidirectional. Also, the combined effect of curvature and pulsatility in cardiovascular flows produces unsteady vortices. In this work, our goal is to assess the correlation between multiple vortex pairs and wall shear stress. To accomplish this, we use an in-house high-order flux reconstruction Navier-Stokes solver to simulate pulsatile flow of a Newtonian blood-analog fluid through a rigid 180$^\circ$ curved artery model. We use a physiologically relevant flow rate and generate results using both fully developed and uniform entrance conditions, the latter motivated by the fact that flow upstream to a curved artery may not be fully developed. Under these two inflow conditions, we characterize the evolution of various vortex pairs and their subsequent effect on several wall shear stress metrics.   

Spiral blood flows in an idealized 180-degree curved artery model

Understanding of cardiovascular flows has been greatly advanced by the Magnetic Resonance Velocimetry (MRV) technique and its potential for three-dimensional velocity encoding in regions of anatomic interest. The MRV experiments were performed on a 180-degree curved artery model using a Newtonian blood analog fluid at the Richard M. Lucas Center at Stanford University employing a 3 Tesla General Electric (Discovery 750 MRI system) whole body scanner with an eight-channel cardiac coil. Analysis in two regions of the model-artery was performed for flow with Womersley number=4.2. In the entrance region (or straight-inlet pipe) the unsteady pressure drop per unit length, in-plane vorticity and wall shear stress for the pulsatile, carotid artery-based flow rate waveform were calculated. Along the 180-degree curved pipe (curvature ratio =1/7) the near-wall vorticity and the stretching of the particle paths in the vorticity field are visualized. The resultant flow behavior in the idealized curved artery model is associated with parameters such as Dean number and Womersley number. Additionally, using length scales corresponding to the axial and secondary flow we attempt to understand the mechanisms leading to the formation of various structures observed during the pulsatile flow cycle.   

Fluid-structure interaction simulations of deformable structures with non-linear thin shell elements

Large deformation of structures in a fluid is simulated using a strongly coupled partitioned fluid-structure interaction (FSI) approach which is stabilized with under-relaxation and the Aitken acceleration technique. The fluid is simulated using a recently developed implicit Newton-Krylov method with a novel analytical Jacobian. Structures are simulated using a triangular thin-shell finite element formulation, which considers only translational degrees of freedom. The thin-shell method is developed on the top of a previously implemented membrane finite element formulation. A sharp interface immersed boundary method is used to handle structures in the fluid domain. The developed FSI framework is validated against two three-dimensional experiments: (1) a flexible aquatic vegetation in the fluid and (2) a heaving flexible panel in fluid. Furthermore, the developed FSI framework is used to simulate tissue heart valves, which involve large deformations and non-linear material properties.   

Hemodynamic Based Coronary Artery Aneurysm Thrombosis Risk Stratification in Kawasaki Disease Patients

Coronary artery aneurysms (CAA) as a result of Kawasaki Disease (KD) put patients at risk for thrombosis and myocardial infarction. Current AHA guidelines recommend CAA diameter \textgreater 8 mm or Z-score \textgreater 10 as the criterion for initiating systemic anticoagulation. Our hypothesis is that hemodynamic data derived from computational blood flow simulations is a better predictor of thrombosis than aneurysm diameter alone. Patient-specific coronary models were constructed from CMRI for a cohort of 10 KD patients (5 confirmed thrombosis cases) and simulations with fluid structure interaction were performed using the stabilized finite element Navier-Stokes solver available in SimVascular. We used a closed-loop lumped parameter network (LPN) to model the heart and vascular boundary conditions coupled numerically to the flow solver. An automated parameter estimation method was used to match LPN values to clinical data for each patient. Hemodynamic data analysis resulted in low correlation between Wall Shear Stress (WSS)/ Particle Residence Time (PRT) and CAA diameter but demonstrates the positive correlation between hemodynamics and adverse patient outcomes. Our results suggest that quantifying WSS and PRT should enable identification of regions at higher risk of thrombosis. We propose a quantitative method to non-invasively assess the abnormal flow in CAA following KD that could potentially improve clinical decision-making regarding anticoagulation therapy.   

Divergence-free smoothing for MRV data on stenosed carotid artery phantom flows

Magnetic Resonance Velocimetry (MRV) is a versatile technique for measuring flow velocity using an MRI machine. It is frequently used for visualization and analysis of blood flows. However, it is difficult to accurately estimate hemodynamics parameters like wall shear stress (WSS) and oscillatory shear index (OSI) due to its low spatial resolution and low signal-to-noise ratio. We suggest a divergence-free smoothing (DFS) method to correct the erroneous velocity vectors obtained with MRV and improve the estimation accuracy of those parameters. Unlike previous studies on DFS for a wall-free flow, we developed a house code to apply a DFS method to a wall-bounded flow. A Hagen-Poiseuille flow and stenosed carotid artery phantom flows were measured with MRV. Each of them was analyzed for validation of the DFS code and confirmation on the accuracy improvement of hemodynamic parameters. We will discuss the effects of DFS on the improvement of the estimation accuracy of velocity vectors, WSS, OSI and etc in detail